Data Obtain and Clean

knitr::opts_chunk$set(echo = TRUE, message = F, warning = F, fig.align = 'center')
suppressPackageStartupMessages(library(tidyverse)) # Data manipulation
library(googlesheets4) # Reading in google sheets
suppressPackageStartupMessages(library(janitor)) # For cleaning
library(naniar) # For visualizing missing data
suppressPackageStartupMessages(library(lares)) # For visualizing missing data
library(readr) # For writing and reading data
library(here) # For good file structure and replication

library(gt) # Nice tables
htmltools::tagList(rmarkdown::html_dependency_font_awesome()) # Needed so fa's in footer will show

teaching_lab_df <- read_sheet("https://docs.google.com/spreadsheets/d/10R_NGttRICpz5uIMjIfcNmYDC4cOkzm1Esl_eDnOKeo/edit#gid=0", 
                              col_types = "Dcccccccccccccccccccdcc") # col_types to specify as D = date, c = character, d = double
# Could easily be done with readxl::read_excel as well

teaching_df <- teaching_lab_df %>%
  rename_with( ~ gsub("\\.\\.\\*$", "", .)) %>% # use regex and dplyr to clean up column names, this one simply deletes everything with 2 or more dots in the column names
  clean_names() # automated column name cleaning with janitor
compare_df_cols(teaching_df, teaching_lab_df, bind_method = "bind_rows") %>%
  gt::gt() %>%# check column name difference
  tab_header(title = md("**Comparing Columns Before and After Janitor**")) %>%
  tab_options(
    table.background.color = "lightcyan"
  )

cols <- c("professional_training_session", 
          "select_your_site_district_parish_or_network", 
          "select_the_best_description_for_your_role",
          "select_the_grade_band_s_you_focused_on", 
          "learning_session_satisfaction",
          "todays_topic_was_relevant_for_my_role",
          "designed_to_help_me_learn",
          "likely_to_apply_46_weeks",
          "select_the_name_of_your_first_facilitator",
          "s_he_facilitated_the_content_clearly_12",
          "did_you_have_a_second_facilitator",
          "select_the_name_of_your_second_facilitator",
          "s_he_facilitated_the_content_clearly_16",
          "s_he_effectively_built_a_community_of_learners_17") # Vector of columns to make factors
cols_agree <- c("learning_session_satisfaction", 
                "todays_topic_was_relevant_for_my_role", 
                "designed_to_help_me_learn", 
                "likely_to_apply_46_weeks",
                "s_he_facilitated_the_content_clearly_12",
                "s_he_facilitated_the_content_clearly_16",
                "s_he_effectively_built_a_community_of_learners_17",
                "s_he_effectively_built_a_community_of_learners_13") # Vector of columns to add levels
teaching_df <- teaching_df %>%
  rename(learning_session_satisfaction = i_am_satisfied_with_the_overall_quality_of_todays_professional_learning_session,
              designed_to_help_me_learn = the_activities_of_todays_session_were_well_designed_to_help_me_learn,
              learning_to_try_out = what_is_the_learning_from_this_professional_learning_that_you_are_most_excited_about_trying_out,
              likely_to_apply_46_weeks = how_likely_are_you_to_apply_this_learning_to_your_practice_in_the_next_4_6_weeks,
              likely_to_recommend_colleague_friend = how_likely_are_you_to_recommend_this_professional_learning_to_a_colleague_or_friend) %>%
  mutate_each_(funs(factor(.)), 
    vars = cols) %>%
  mutate_each_(funs(factor(.,
                           levels = c("Strongly disagree", "Disagree", "Neither agree nor disagree", "Agree", "Strongly agree"))),
               vars = cols_agree) # First rename some of the columns then make them factors and add levels, cols filters which variables to do this for 

original_columns <- colnames(teaching_lab_df) # Just useful to have for later, not important if knit

# write_rds(teaching_df, here("Data/final_df.rds"))
# write_rds(teaching_lab_df, here("Data/original_df.rds"))